Long-Proboscid Brachyceran Flies in Cretaceous Amber (Diptera

SYEN syen_12106 Dispatch: September 15, 2014 Journal: SYEN CE: Journal Name Manuscript No. B Author Received: No of pages: 26 TS: Naresh

Systematic Entomology (2014), 0, 0–0 DOI: 10.1111/syen.12106

1 1 2 2 3 3 4 4 5 Long-proboscid brachyceran ﬂies in Cretaceous amber 5 6 6 7 (Diptera: Stratiomyomorpha: Zhangsolvidae) 7 8 8 9 9 1 2 10 ANTONIO ARILLO , ENRIQUE PEÑALVER , RICARDO PÉREZ 10 11 -DELAFUENTE3, XAVIER DELCLÒS4, JULIA CRISCIONE5, 11 12 PHILLIP M. BARDEN5, MARK L. RICCIO6 and D AV I D A . 12 13 5 13 14 GRIMALDI 14 15 1Departamento de Zoología y Antropología Física, Facultad de Biología, Universidad Complutense, Madrid, Spain, 2Museo 15 16 Geominero, Instituto Geológico y Minero de España, Madrid, Spain, 3Museum of Comparative Zoology, Harvard University, 16 17 4 17 AQ1 Cambridge, MA, U.S.A., Departament d’Estratigrafia, Paleontologia i Geociències Marines, Facultat de Geologia, Universitat de 18 Barcelona, Barcelona, Spain, 5Division of Invertebrate Zoology, American Museum of Natural History, New York, NY, U.S.A. and 18 19 6Institute of Biotechnology, Cornell University, Ithaca, NY, U.S.A. 19 20 20 21 21 22 Abstract. The monophyletic family Zhangsolvidae comprises stout-bodied brachyc- 22 23 eran flies with a long proboscis and occurring only in the Cretaceous, originally known 23 24 in shale from the Early Cretaceous Laiyang Formation (Fm.) in China (Zhangsolva 24 25 Nagatomi & Yang), subsequently from limestones of the Early Cretaceous Crato Fm. 25 26 of Brazil. Cratomyoides Wilkommen is synonymized with Cratomyia Mazzarolo & 26 27 Amorim, both from the Crato Fm.; Cratomyiidae is synonymized with Zhangsolvidae. 27 28 28 Two genera and three species of Zhangsolvidae are described: Buccinatormyia magnifica 29 29 30 Arillo, Peñalver & Pérez-de la Fuente, gen. et sp.n. and B. soplaensis Arillo, Peñalver & 30 31 Pérez-de la Fuente, sp.n., in Albian amber from Las Peñosas Fm. in Spain; and Lingua- 31 32 tormyia teletacta Grimaldi, gen. et sp.n., in Upper Albian–Lower Cenomanian amber 32 33 from Hukawng Valley in Myanmar. Buccinatormyia soplaensis and Linguatormyia tele- 33 34 tacta are unique among all Brachycera, extant or extinct, by their remarkably long, 34 35 flagellate antennae, about 1.6× the body length in the latter species. A phylogenetic anal- 35 36 ysis of 52 morphological characters for 35 taxa is presented, 11 taxa being Cretaceous 36 37 species, which supports placement of the family within Stratiomyomorpha, although not 37 38 to any particular family within the infraorder. 38 39 39 40 This published work has been registered in Zoobank, http://zoobank.org/urn:lsid: 40 41 zoobank.org:pub:F32CF887-7C37-45D5-BD6B-135FE9B729A7. 41 42 42 43 43 44 44 45 Introduction Formation limestone of Brazil (Mazzarolo & Amorim, 2000; 45 46 Wilkommen & Grimaldi, 2007). The genus Cratomyia was 46 47 The Zhangsolvidae is an extinct family of brachyceran flies placed in the infraorder Stratiomyomorpha by Mazzarolo & 47 48 erected by Nagatomi & Yang (1998) for Zhangsolva cupressa Amorim (2000), a basal lineage of Brachycera that includes 48 49 (Zhang, Zhang & Li), preserved in shale from the Early three living families: the Stratiomyidae Latreille, a cosmopoli- 49 50 Cretaceous of China. The family was subsequently found to tan group comprising 2651 species in 375 genera (Woodley, 50 51 include two species of Cratomyia from the Aptian-aged Crato 2001); the Xylomyidae Verrall, a small cosmopolitan fam- 51 52 ily comprising approximately 138 species in four genera 52 53 Correspondence: Ricardo Pérez-de la Fuente, Museum of Compara- (Woodley, 2011); and the Pantophthalmidae Bigot, a small 53 54 tive Zoology, Harvard University, 26 Oxford St., Cambridge, MA 02138, neotropical family of very large flies with 20 species in two 54 55 U.S.A. E-mail: [email protected] genera (Val, 1976). Several other late Mesozoic families have 55 56 been suggested to be within this infraorder, but their rela- 56 57 The authors declare no conflict of interest. tionships remain poorly understood. These families include 57

1 Eremochaetidae Ussatchov, Tethepomyiidae Grimaldi & Arillo purchased in 2012 from Mr. Sieghard Ellenberger, Germany, 1 2 and, perhaps, Oligophrynidae Rohdendorf (Krzeminski´ & who obtained the amber from mines near Myitkyina, Kachin 2 3 Krzeminska,´ 2003; Grimaldi & Arillo, 2008), although the Province, northern Myanmar (see Grimaldi et al., 2002). The 3 4 focus of this paper is on the position of Zhangsolvidae. amber surfaces lying over the dorsal and ventral portions of the 4 5 In this paper we describe four new Cretaceous specimens fly were slightly flattened using a lapidary wheel; there wasno 5 6 of Zhangsolvidae, three of them originally found in the same other preparation. The amber piece is occluded with a turbid sus- 6 7 amber piece from the Spanish locality of El Soplao, plus one pension of particles, so in order to visualize critical details it 7 8 specimen from Burmese amber. The El Soplao outcrop, Albian was CT scanned at the Cornell Biotechnology Resource Center 8 9 in age, was recently discovered in Spain (Najarro et al., 2009, Imaging Facility. 9 10 2010; Peñalver & Delclòs, 2010). To date, insects belonging The Burmese specimen scanning was performed using a Zeiss 10 11 to 13 orders and about 40 families are preserved in this amber VERSA XRM-520 instrument. For each dataset, 2400 X-Ray 11 12 (Pérez-de la Fuente, 2012). In the order Diptera, the families projections were obtained at 0.15∘ intervals over 360∘ using 12 13 Atelestidae, Cecidomyiidae, Ceratopogonidae, Chimeromyi- 60 kV, 32 μA and 5000 ms exposure time, with a 2048 × 2048 13 14 idae, Chironomidae, Dolichopodidae, Hybotidae, Phoridae, pixel detector using a 1.5 μm pixel size that was 2 × 2 binned into 14 15 Psychodidae and Rhagionidae have been found (ibid.), although an effective pixel size of 3.0 μm. Proper placement of the speci- 15 16 only ceratopogonids have been studied (Pérez-de la Fuente men was accomplished with the assistance of X-Ray fluoroscopy 16 17 et al., 2011). mode, to ensure that the area of interest was included within the 17 18 Burmese amber contains probably the most diverse palaeo- scanning field of view. The projection views were used to recon- 18 19 biota of the seven major world deposits of amber from the Cre- struct a CT image using a convolution back-projection approach 19 20 taceous Period, with approximately 228 families of organisms implemented in 3D, giving a 3 × 3 × 3 mm volume of image data 20 21 (primarily arthropods) thus far reported, Diptera being espe- with 3 μm isotropic voxels in arbitrary density units (ADU) and 21 22 cially diverse (Zherikhin & Ross, 2000; Grimaldi et al., 2002; exported as DICOM or TIFF stacks. Reconstructed data files 22 23 Ross et al., 2010). Burmese amber contains, for instance, the were imported into the software programs OsiriX (64 bit, v5.5) 23 24 oldest definitive Mesozoic record of mosquitoes (Culicidae) and Avizo 8.0 (VSG), where they were visualized and analysed 24 25 (Borkent & Grimaldi, 2004). Estimated age of ‘burmite’ was using a variety of 2D and 3D techniques. 25 26 recently established as the Albian–Cenomanian boundary, c. Fly morphological terminology generally follows McAlpine 26 27 99 Ma, based on radiometric dating of zircons from the amber (1981). We do not use the term postpedicel, used by some dipter- 27 28 matrix (Shi et al., 2012). ists to refer to that portion of the antenna distal to the pedicel, 28 29 The new zhangsolvids are the first specimens preserved as because by definition this is the flagellum. Szucsich & Krenn 29 30 amber inclusions and the youngest occurrences in the fossil (2000, 2002) introduced the term ventral rostral membrane, 30 31 record of the family, significantly expanding its palaeobiogeo- which we use here, referring to the pleated membrane at the base 31 32 graphic range. The exquisite preservation in amber allows obser- of the proboscis. 32 33 vation of minute morphological details, providing substantially Although the three Spanish amber specimens are syninclu- 33 34 more data to test the proposed relationships of Zhangsolvidae sions, the gender of only one of them is definitive. Because 34 35 within the Stratiomyomorpha. These findings also provide sig- another specimen is substantially larger but lacks terminalia 35 36 nificant palaeoecological data, but these are presented elsewhere (and has significant differences in venation and antennal struc- 36 37 (E. Peñalver et al., in preparation). ture), the lengths of the thorax and wing in 34 exemplar extant 37 38 species of Stratiomyidae, Xylomyidae and Pantophthalmidae 38 39 were measured for at least one male and one female, to test 39 40 Material and methods for the presence/extent of sexual dimorphism in body size. We 40 41 used pinned specimens from the AMNH collection for these 41 42 The Spanish specimens were discovered in a large piece of measurements. For the first two of these families the measure- 42 43 amber containing numerous inclusions from the El Soplao out- ments were made using a Nikon SMZ1500 stereomicroscope 43 44 crop, near the municipality of Rábago (Autonomous Commu- and NIS Elements software; measurements of Pantophthalmi- 44 45 nity of Cantabria). This piece was trimmed into several smaller dae were made using a millimetre ruler because of their large 45 46 pieces to better allow examination of individual inclusions. sizes. 46 47 Each of these smaller pieces was embedded in synthetic resin Phylogenetic analysis used morphological characters from 47 48 (EPO-TEK 301) and then polished (see Nascimbene & Silver- exemplar extant and fossil stratiomyomorphans, as well as 48 49AQ2 stein, 2000). Fossils were examined with an Olympus BX51 all species of Zhangsolvidae, analysed under equal-weights 49 50 compound microscope. Photomicrographs were taken using an parsimony with the program TNT (Goloboff et al., 2008). 50 51 attached camera, some of them z-stacked. Drawings were made Parameters employed an ‘XMULT’ search, which uses sectorial 51 52 using an Olympus U-DA drawing tube attached to the com- searching, tree drifting, ratcheting and fusing. The search ran 52 53 pound microscope. The Spanish specimens are provisionally until the optimal tree score was found 100× independently, 53 54 housed at the Museo Geominero (IGME) in Madrid (Spain); whereupon all lowest-cost topologies were consolidated into 54 55 their final deposit will be in the collection of the El Soplao Cave a strict consensus. Consistency and retention indices were 55 56 (Government of Cantabria). The Burmese amber specimen is in calculated, along with Bremer and bootstrap support values 56 57 the American Museum of Natural History, New York, and was (with 10 000 pseudoreplicates). 57

1 Systematic palaeontology 12-segmented, 1 2 (iii) proboscis conspicuously long and (iv) costal cell concave 2 3 Order Diptera Linnaeus, 1758 before Rs, possibly representing a sc-r crossvein. The holotype 3 4 Infraorder Stratiomyomorpha Hennig, 1973 was not studied by Nagatomi & Yang, so they assumed that the 4 5 original illustrations were accurately drawn. Unfortunately, the 5 6 6 AQ3 Family Zhangsolvidae Nagatomi & Yang, 1998 holotype of Z. cupressa (a unique specimen) is currently lost (Dr 7 7 Cratomyiidae Mazzarolo & Amorim, 2000: 94. Type genus and H.-C. Zhang, personal communication to A.A., 2011), which is 8 8 species: Cratomyia macrorrhyncha Mazzarolo & Amorim (by why we were unable to restudy the holotype. According to the 9 9 original designation). NEW SYNONYMY. ICZN code, dramatic taxonomic changes would be necessary 10 for the designation of a new type specimen of the family (from 10 11 a different type genus), which we have not made in case the lost 11 12 Type genus and species. Zhangsolva Nagatomi & Yang, specimen eventually appears in a Chinese collection, or a new 12 13 1998. Holotype of the type species Archisolva cupressa Zhang, specimen (and potential neotype) is excavated from the Laiyang 13 14 Zhang & Li, 1993, is currently lost (Dr H.-C. Zhang, personal Formation. 14 15 communication to A.A., 2011). Archisolva was preoccupied in Later, Mazzarolo & Amorim (2000) described Cratomyia 15 16 Stratiomyidae, so Nagatomi &Yang 1998 provided the replace- macrorrhyncha, from the Early Cretaceous Crato Formation, 16 17 ment generic name Zhangsolva. Brazil, for which they proposed the family Cratomyiidae. Their 17 18 study revealed that Cratomyia was a sister group to (Xylomyi- 18 19 Taxonomic composition. Zhangsolva cupressa (Zhang et al., dae + Stratiomyidae), but these authors had not taken into 19 20 1993); Cratomyia macrorrhyncha Mazzarolo & Amorim, 2000; account Zhangsolva cupressa. The most recent contribution on 20 21 21 AQ4 Cratomyia cretacica (Wilkommen, 2007), comb.n.; Buccina- this group of flies was the description of Cratomyoides cretaci- 22 tormyia magnifica Arillo, Peñalver & Pérez-de la Fuente, gen. cus Wilkommen (in Wilkommen & Grimaldi, 2007), a fossil fly 22 23 et sp.n.; Buccinatormyia soplaensis Arillo, Peñalver & Pérez-de also found in the Crato Formation and very similar to Cratomyia. 23 24 la Fuente, sp.n.; Linguatormyia teletacta Grimaldi gen. et sp.n. In that paper the authors recognized the close relationship among 24 25 Zhangsolva, Cratomyia and Cratomyoides. Recently, Grimaldi 25 26 et al. (2011) suggested that all of these fossils should be included 26 Diagnosis (emended). Proboscis rather long to very long 27 in the family Zhangsolvidae (with Cratomyiidae as a junior syn- 27 (from 1 to about 4× the head length), slender, apex with small 28 onym). They also suggested that Cratomyoides is a junior syn- 28 labellum; vein C ending at or near apex of wing, M strongly 29 1 onym of Cratomyia, but they did not formally synonymize these 29 arched, M fused to CuA and CuA fused to A near wing 30 3 1 2 1 taxa. In the same work the authors also indicated that several 30 margin. 31 features in the original description of Zhangsolva were probably 31 32 erroneous: (i) parts of the wing venation are obvious artefacts, 32 33 Remarks. We have refrained from including the following such as a sc-r crossvein, which is present in only one wing (as 33 34 characters in the family diagnosis above, as they are not known seen in the photograph of the holotype in the original paper), 34 35 from all the zhangsolvid genera (only from half or less of the and (ii) the antennae probably have no more than eight flagel- 35 36 species, mainly the ones preserved in amber), but we posit that lomeres (the groundplan in Brachycera), a misinterpretation per- 36 37 they could be of importance to characterize the family: (i) maxil- haps originating from preservation (e.g. transverse cracks in car- 37 38 lary palp two-segmented, long and projecting (observed in Buc- bonaceous films commonly occur in compression-fossil insects 38 39 cinatormyia magnifica gen. et sp.n.andLinguatormyia teletacta preserved in rock). Furthermore, the long basal section of Rs 39 40 gen. et sp.n., tentatively observed in Cratomyia macrorrhyncha, proposed as diagnostic for Zhangsolvidae for Nagatomi & Yang 40 41 see Mazzarolo & Amorim, 2000), (ii) tibial spurs 0, 2, 2 (only (1998) is quite short in the new Burmese amber species. Hence, 41 42 observed in B. magnifica), and (iii) pulvilli and empodium well the conspicuously long proboscis is the only character that we 42 43 developed, pulvilliform (idem). Some zhangsolvids, i.e. Bucci- have retained here from the original diagnosis of the family. 43 44 natormyia soplaensis sp.n. and L. teletacta, are readily recog- 44 45 nizable by showing remarkable flagellate antennae that can be 45 Genus Cratomyia Mazzarolo & Amorim 46 more than 1.5× longer than the body (in the latter species). 46 Cratomyia 47 The species Archisolva cupressa was described by Zhang Mazzarolo & Amorim, 2000: 94. Type species: 47 C. macrorrhyncha 48 et al. (1993) from the Early Cretaceous Laiyang Formation Mazzarolo & Amorim. By original 48 49 in Laiyang, Shandong Province, China, originally placed in designation. 49 Cratomyoides 50 the family Solvidae (=Xylomyidae). Later, Nagatomi & Yang Wilkommen, 2007, in Wilkommen & Grimaldi, 50 Cratomyoides cretacicus 51 (1998) proposed a replacement name, as the generic name 2007: 375. Type species: Wilkommen. 51 52 was already occupied by Archisolva Enderlein (Stratiomyi- By original designation. NEW SYNONYMY. 52 53 dae). They provided the new name Zhangsolva and considered 53 54 that, even though the genus was close to Stratiomyoidea, it Diagnosis (emended). Cratomyia is readily separated from 54 55 represented a new family. Diagnostic characters of the new other (Laurasian) genera of Zhangsolvidae by the unique pres- 55

56 family Zhangsolvidae proposed by Nagatomi & Yang (1998) ence in the former of a stem to veins M1 and M2 distal to cell d 56 57 were: (i) long basal section of Rs, (ii) antennal flagellum (instead of diverging directly from cell d). 57

1 1 Genus Buccinatormyia Arillo, Peñalver & Pérez-de la level of r-m crossvein, cell m3 narrow (W/L = 0.21), M2 and M3 2 Fuente, gen. n nearly parallel, CuP well sclerotized. 2 3 http://zoobank.org/urn:lsid:zoobank.org:act:D45D7DD7- 3 4 4 89E2-450F-BA5E-46F92F66611A ♀ 5 Description ( ; based largely on the holotype except for wing 5 6 features entirely based on the paratype). Body 9.20 mm long 6 Type species. Buccinatormyia magnifica sp.n. 7 (estimated from reconstruction). 7 8 Head: Broad, 1.82 mm wide. Eyes large, bulging, occupying 8 9 Etymology. Derived from buccinator- (L.), trumpeter, and nearly entire lateral surface of head (genae exposed ventrally), 9 10 -myia (Gr.), fly, in reference to the long proboscis. widely separated, ventrally 0.73 mm wide between inner mar- 10 11 gins; bare, devoid of interfacetal setulae; facets undifferenti- 11 ated; inner margins in frontal view without emargination around 12 Diagnosis. Antenna with six articles (four flagellomeres), 12 antennal base. Eyes significantly broader dorsally. Dense, long, 13 with minute apical stylus; stem of Rs long (several times the r-m 13 14 fine, stiff setae around eye margin. Three small ocelli present, 14 crossvein length), C vein ending beyond R5 but without reaching 15 lying in a clearly demarcated, densely setose ocellar triangle. 15 M1,veinsM1 and M2 diverging directly from cell d (not from a 16 Vertex of head with paramedian emarginations, centrally raised 16 portion of M1+2 distad to cell d). 17 into median dorsal tubercle; vertex with three ranges of pilos- 17 18 ity: (i) vestiture of fine, dense setulae as on the rest of the head, 18 Buccinatormyia magnifica Arillo, Peñalver & Pérez-de la 19 (ii) sparse, longer setae, and (iii) long, fine, stiff setae like those 19 Fuente, sp. n 20 around eye margin. Areas surrounding antennal bases raised into 20 21 http://zoobank.org/urn:lsid:zoobank.org:act:21A4E7D7-D8F6- bare, rounded mounds (=facial lobes), nearly meeting medi- 21 22 4FBB-9736-73F5F80F6066 ally. Front of head with well-developed, inverted Y-shaped sul- 22 23 (Figs 1–8) cus. Clypeus large, bulging, covered by fine, dense setulae, 23 24 raised into low mound between facial lobes and palpal bases, 24 25 Type material. Holotype CES-349.1 and paratype in two fully frontal in position; with short vertical and median sulci 25 26 parts, CES-015.1 and CES-392.3, from El Soplao amber and a transverse, median depression; ventral surface of clypeus 26 27 (Spain). These specimens and holotype of B. soplaensis sp.n. membranous, oval; membrane connected to base of proboscis. 27 28 were found in the same amber piece as syninclusions, which Frontoclypeal sulci (demarcating lateral edges of clypeus) very 28 29 was prepared in three sample pieces for study (CES-349: deep. Genae developed into pair of rounded, ventral lobes. Neck 29 30 amber fragment 10 × 5 × 5mminaresinprism16× 8 × 5 mm; entirely membranous. 30 31 CES-015: amber fragment 13 × 11 × 6 mm in a resin trapezoid Proboscis: Long (virtually all of it haustellum), 3.85 mm 31 32 15 × 19 × 6 mm; and CES-392: amber fragment 24 × 10 × 7mm long from base to apex (0.42× the body length), 0.05 mm 32 33 in a resin prism 30 × 11 × 8 mm). Holotype is in a good state wide, 0.17 mm high at middle. Proboscis laterally flattened, 33 34 of preservation, virtually complete, but wings are incomplete 0.05 mm wide, 0.17 mm high at the middle (width 0.31× the 34 35 and a dorsal portion of the thorax is slightly altered. Also, cross-sectional depth) and directed forward in relation to head. 35 36 proboscis, legs and abdomen are crossed by a fracture that par- Proboscis inserted below facial margin. Base of proboscis ven- 36 37 tially destroyed some small portions of the abdomen. Paratype trally with annulated membrane (ventral rostral membrane), 37 38 is strongly compressed but virtually complete, lacking only five-folds visible, each with numerous annulations (membrane 38 39 the distal portions of the proboscis and the abdomen; it lies in probably telescoped). Proboscis covered by dense microtrichia 39 40 two prepared sample pieces, the smaller portion (CES-392.3) and a dense vestiture of large, stiff deeply-ribbed setae with stel- 40 41 consisting of a costal fragment of wing with the halter, several late transverse section and directed forward. Haustellum exter- 41 42 leg portions and a small portion of the abdominal cuticle. nally composed by a tube-like structure with a dorsal groove 42 43 (=theca), subellipsoidal in cross-section and with straight lat- 43 44 Syninclusions. Holotype and paratype of the new species eral walls (well preserved in holotype; laterally collapsed 44 45 and the holotype of B. soplaensis sp.n. were preserved in one in paratype, ventrally forming an artifactual surface). Dorsal 45 46 large piece of amber, which was divided into several pieces to groove internally closed by a narrow (0.05-mm wide), semi- 46 47 allow study of the 48 arthropods in it, including ten additional transparent membrane with thickened margins and covered dor- 47 48 dipterans, mainly small brachycerans. sally by a fine vestiture of minute setae, interpreted as a modified 48 49 labrum (most visible in the basal half of the holotype’s pro- 49 50 boscis, partially broken and protruding due to downward bend- 50 51 Etymology. Named after the wonderful state of preservation ing of the proboscis). Labrum externally covering the proboscis 51 52 and the impressive habitus of the two specimens. base and basally entering into the proboscis. Epi/hypopharynx 52 53 not visible, but most likely forming floor of proboscis. Food 53 54 Diagnosis. Proboscis very elongate (about 0.40× the body canal preserved, 0.03 mm wide and 0.07 mm high as preserved) 54 55 length); antenna not flagellate and short (about× 0.30 the (visible in the amber surface that cross-sections the paratype’s 55 56 wing length), flagellomere II with dorsal surface concave, proboscis). Labellum very small, about as wide as proboscis, 56 57 flagellomere III about 0.30× the antennal length; Sc ending at oval in cross-section. Labellar lobes closely pressed together 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 Fig. 1. Buccinatormyia magnifica gen. et sp.n., from El Soplao amber. Entire holotype CES-349.1 (ventro-oblique view), shown as preserved. 49 50 50 51 51 52 medially, 0.22 mm long, 0.13 mm high; each lobe has around 13 basally and gradually narrowing distally, 0.15 mm high at the 52 53 pseudotracheae with transverse pseudotracheal rings on exposed base (greatest height), 0.06 mm high at apex (lowest height); 53 54 surfaces. distal 2/3 of its length laterally flattened, 0.11 mm wide at the 54 55 Maxillary palp: Elongate, 1.45 mm long; projected forward; base (greatest width). Basal palpomere with abundant, very long, 55 56 two-segmented. Bases of palps flank base of proboscis. Basal fine, stiff setae along its length, oriented at90∘-angle to the 56 57 palpomere straight and long, 0.60× the total palp length; inflated antennal axis. Apical palpomere long, club-like, pedunculate, 57

1 comprising dense, small foveae (probably sensilla much smaller 1 2 than other foveae on antenna and palp). Flagellomere IV acumi- 2 3 nate distally. Antenna with minute apical stylus. 3 4 Thorax: Narrow and deep as preserved, 2.34 mm long, 4 5 1.20 mm wide (estimated), 1.88 mm high, slightly narrower 5 6 than head (thoracic cuticle of holotype partially lost, mainly 6 7 anterodorsally; thorax of paratype slightly collapsed, pleuron 7 8 not well seen). Thorax devoid of macrosetae, exocuticle with 8 9 fine, short setae only; coloration appears uniformly dark brown- 9 10 ish (possibly preservational). Anterior part of thorax obscure. 10

11 Color Mesothorax large, expansive. Katepisternum ventrally broad 11 12 and slightly bulging. Anepisternum large; anepimeron smaller, 12 13 subdivided by partial cleft; mesothoracic spiracle not visible; 13 14 meron slightly larger than mesocoxa. Coxa and ventral surfaces 14 15 of femur with sparse, short setulae. Procoxa slender, 1.02 mm 15 16 long; mesocoxa 0.64 mm long, 0.23 mm wide basally; metacoxa 16 17 0.60 mm long. Anterodistal surfaces of each coxa with small, 17 18 protruding lobe. Trochanter small, slender, fully articulated 18 19 with femur and coxa (no fusion). Legs slender. Prothoracic 19 20 leg about as long as other legs: femur 1.58 mm long, tibia 20 21 1.63 mm long and tarsomere I 2.38 mm long. Mesothoracic 21 22 leg 5.85 mm long, as estimated: femur 1.70 mm long as esti- 22 23 mated, tibia 1.95 mm long, tarsomere (ta) I 1.22 mm long, taII 23 24 0.45 mm long, taIII mm 0.23 long, taIV 0.13 mm long and taV 24 25 0.18 mm long. Metathoracic leg 6.83 mm long: femur 2.03 mm 25 26 long, tibia 2.50 mm long, tarsomere (ta) I 1.30 mm long, taII 26 27 0.48 mm long, taIII 0.25 mm long, taIV 0.13 mm long and taV 27 28 0.15 mm long. Femora with recumbent, dense setulae. Tibiae 28 29 with denser, stronger setae. Protibia without spurs. One pair of 29 30 Fig. 2. Buccinatormyia magnifica gen. et sp.n., from El Soplao amber. setose distal spurs each on meso- and metatibia. Tarsomeres 30 31 Photomicrograph of the holotype, CES-349.1. covered with abundant, short, strong setae. Tarsi more setose 31 32 than tibiae, especially strong setae distally and each tarsomere 32 33 strongly laterally flattened, very broad distally, apex rounded; having several distal, spine-shaped setae. Claws covered by 33 34 with several fine foveae (probably sensilla) distributed mainly microtrichia. Pulvilli paddle-shaped; flat, paddle-shaped (pul- 34 35 distally but without apparent pattern. villiform) empodia on all pretarsi, of similar size to pulvilli; all 35 36 Antenna: Six-articled (flagellum with four flagellomeres), pro- three pads covered by microtrichia. 36 37 jecting forward; 1.85 mm long (0.20× estimated body length). Wing (based on paratype, including halter): Length 6.00 mm, 37 38 Antennal bases narrowly separated by distance subequal to 2.09 mm wide (greatest width) (W/L = 0.35). Costal vein 38

39 diameter of antennal socket (c. 0.80× distance between palp ending between R5 and M1. Veins Sc and R fused basally to 39 40 bases). Scape short, c. twice as long as wide, 0.18 mm long, crossvein h; Sc 2.57 mm long (0.43× the wing length). Vein 40 41 0.10 mm wide, 0.14 mm high; both scapes oriented antero- Sc meets wing margin beyond middle of wing length, at level 41

42 laterally (divergent) as preserved. Scape and pedicel covered of vein r-m. Vein R1 3.15 mm long, gradually thickened api- 42 43 by sparse, relatively strong setae projected forward. Antenna cally; stem of Rs relatively long, 0.89 mm long (ca. 4.5× the 43

44 entirely covered by a vestiture of dense, fine setulae. Pedicel length of r-m crossvein); R2+3 1.62 mm long; stem of R4+5 44 45 short, 0.75× the scape length, as wide as scape. Flagellum rel- 1.51 mm long, R4 and R5 short, asymmetrical, R4 0.57 mm 45 46 atively short, laterally flattened, comprising four flagellomeres: long, R5 1.05 mm long; tip of R5 preapical to wing tip. Cell 46 47 I 0.27 mm long, 0.06 mm wide, 0.09 mm high, approximately d short and thick, 1.03 mm long (greatest length), 0.40 mm 47

48 as long as scape + pedicel; II 0.56 mm long, 0.05 mm wide, wide (W/L = 0.39). Stem of M1+2 0.86 mm long; M1 arched 48 49 0.09 mm high (greatest height); III as long as flagellomere II but costad, tip of M1 ending posterior to wing tip; M1 1.18 mm 49 50 with different morphology, 0.53 mm long, 0.12 mm wide (great- long (linear, not along curvature), M2 0.74 mm long; M1 and 50 51 est width), 0.03 mm high; IV short, as long as scape, 0.18 mm M2 diverge directly from cell d, not from a portion of M1+2 51 52 long, 0.03 mm wide, 0.06 mm high (greatest height). Flagellom- distad to cell d; veins M2 and M3 nearly parallel. M2 ends 52 53 ere II with dorsal surface concave; II and III inserted ventrally before wing margin. Cell m3 long and slender, 1.02 mm long, 53 54 to their proximal flagellomere. Flagellomere III proximally with 0.21 mm wide (W/L = 0.21); veins CuA1 and M3 joined in 54 55 a few fine foveae (probably sensilla), similar to those on distal very short vein before meeting wing margin. Cell br slightly 55 56 palpomere, distribution without apparent pattern; this flagellom- narrower than bm; cell bm slightly longer (2.72 mm long, vs. 56 57 ere with narrow band on ventrolateral surface (except on apex) 2.54 mm). Vein CuA with base approximately half the thickness 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 Fig. 3. Spatial relation of the paratype of Buccinatormyia magnifica gen. et sp.n., the holotype of B. soplaensis sp.n., and some other close 34 35 syninclusions integrating drawings from two preparations of the same amber piece (CES-015 and CES-392). (1) Buccinatormyia soplaensis partially 35 36 reconstructed; (2) Buccinatormyia magnifica; (3–4) Diptera: Brachycera; (5) Neuroptera: Coniopterygidae; (6) Lepidoptera: Micropterigidae; (7) Acari: 36 37 Actinedida. 37 38 38 39 39 of base of R; CuA2 fusing with A1 prior to wing margin, thorax; 5.30 mm long (without cerci); strongly flattened 40 forming closed, narrow cell cup; cup 2.97 mm long dorsoventrally, especially on margins. Abdomen compris- 40 41 41 (W/L = 0.12); CuP well developed and sclerotized, parallel ing eight visible segments; dorsal surface (tergites) flat, 42 42 to CuA and separated by distance equal to vein thickness. virtually without relief. Coloration appears uniformly dark 43 43 Vein CuA strongly curved before connection with A .Vein brownish (possibly preservational). Sternites cover entire trans- 44 2 1 44 A entirely straight; A preserved, basal sclerotized portion verse width of abdomen; no abdominal pleural membrane 45 1 2 45 0.95 mm long. Apices of veins CuA + M and CuA + A exposed. Segments V–VIII laterally form very thin planar 46 1 3 2 1 46 47 very thin, faint. Anal lobe, alula and calypter well developed; expansions (=‘lateral flanges’ sensu Stuckenberg (2000) for 47 48 calypter 0.52 mm long, with marginal vein. Wing membrane Vermileonidae), but very small on segment VIII; segment 48 49 entirely hyaline (no patterning or pterostigmatic sclerotization), IV apparently with a small distal section as lateral flanges. 49 50 evenly covered with minute microtrichia. Halter 1.00 mm long, Abdominal cuticle showing a vestiture constituted by two 50 51 with slender apical knob 0.41 mm long, 0.21 mm wide. Stem different types of specialized setae: (i) ‘stellate hairs’, cor- 51 52 and knob covered by dense microtrichia and knob with a few responding to microtrichia arranged in bunches of c. five, 52 53 additional longer setae. Halter with sensory area at least in present on dorsal surface of lateral flanges and also laterally 53 54 middle of its stem, having campaniform-like sensilla; pilosity at least on segment III, and (ii) large, stiff deeply-ribbed 54 55 of base obscure. setae (stellate in transverse section), similar to those on pro- 55 56 Abdomen (virtually complete in holotype): Broad, broader boscis, sparsely covering the entire abdominal cuticle. Only 56 57 distally, 1.45 mm wide (greatest width), slightly broader than larger lateral flanges (segments V–VII) have dorso-marginal, 57

1 Diagnosis. Proboscis elongate (about 0.30× the body length, 1 2 estimated); antenna flagellate and elongate (about× 0.60 the 2 3 wing length), flagellomere II with dorsal surface straight, flag- 3 4 ellomere III about 0.65× the antennal length; Sc ending beyond 4 5 5 level of r-m crossvein, cell m3 broad (W/L = 0.34), M2 and M3 6 highly divergent, CuP weakly developed. 6 7 7 8 8 9 Description (sex unknown). Body about 14 mm long (esti- 9 10 mate based on comparison of wing lengths between this species 10 B. magnifica 11 Color and ). 11 12 Head: Incomplete (only a portion of eye preserved as 12 13 an impression, showing that eyes are large in this species, 13 14 though insufficiently preserved to determine if holoptic or 14 15 dichoptic). 15 16 Proboscis: Incomplete, preserved portion 1.07 mm long (prox- 16 17 imal end obliquely sectioned by surface of the preparation, 17 18 estimated to have originally been similar or slightly greater in 18 19 length than that of B. magnifica). Proboscis laterally flattened, 19 20 0.22 mm high, directed forward in relation to head. Proboscis 20 21 covered by dense microtrichia and a dense vestiture of large, 21 22 stout setae with deep longitudinal ribbing (stellate in transverse 22 23 section) and directed forward. Haustellum externally composed 23 24 by a tube-like structure with a dorsal groove (=theca), subel- 24 25 lipsoidal in cross-section and with straight lateral sides. Labrum 25 26 Fig. 4. Photomicrograph of the paratype of Buccinatormyia magnifica narrow, 0.06 mm wide, with thickened margins (visible by cutic- 26 27 gen. et sp.n. and the wing of the holotype of B. soplaensis sp.n., ular transparency in a dorsal, slightly oblique position and also 27 28 CES-015. as an oblique section created by the surface of the amber prepa- 28 29 ration). Food canal not visible. Labellum very small, narrower 29 30 longitudinal, curved lines comprising minute ellipsoid plaques than proboscis (preserved in fine detail, located very close to 30 31 of variable size (=abdominal muscle plaques, see Stoffolano preparation surface, thus observable under high magnification). 31 32 et al., 1988). Female terminalia preserved: cercus strongly Labellar lobes cushion-like, 0.26 mm long, c. 0.16 mm wide, c. 32 33 flattened dorsoventrally, two-segmented, covered by fine setae; 0.08 mm high; each lobe has 10–12 pseudotracheae with 30–40 33 34 basal segment 0.21 mm long, 0.08 mm wide; distal segment transverse pseudotracheal rings (one labellar lobe well exposed, 34 35 0.15 mm long, 0.04 mm wide. slightly oblique; other in perpendicular position, obscured to 35 36 observation). Labellar lobes with sparse, strong setae irregularly 36 37 Buccinatormyia soplaensis Arillo, Peñalver & Pérez-de la distributed, as figured. 37 38 Fuente, sp.n. Antenna: Six-articled (flagellum with four flagellomeres), 38 39 http://zoobank.org/urn:lsid:zoobank.org:act:C7BDF2D4-F83E- very elongate and flagellate, projecting forward, with same 39 40 4BC4-AAC0-F7562F656A4A configuration as in B. magnifica sp.n. but with two flagellom- 40 41 (Figs3,4,8,9) eres elongated (c. 2× for flagellomere II and c.7× for flag- 41 42 ellomere III). Scapes not preserved, only apices of pedicels 42 43 preserved; scape + pedicel of same length as in B. magnifica 43 Type material. Holotype in two fragments CES-015.2 and 44 (inferred by the close position of the eye remains); antenna 44 CES-392.2, from El Soplao amber (Spain). Specimen preserved 45 5.62 mm long (estimated from reconstruction). Preserved por- 45 as a complete wing with a small portion of the thoracic cuticle 46 tion of pedicel covered by long setae projecting forward. Flag- 46 (CES-015.2) and distal parts of the antennae and proboscis 47 ellum covered by vestiture of dense, fine setulae. Flagellum 47 including traces of the eyes (CES-392.2). 48 long, laterally flattened, comprising four articles: flagellom- 48 49 ere I 0.34 mm long, 0.03 mm wide, 0.12 mm high; II 1.18 mm 49 50 Syninclusions. The holotype of the new species and the long, 0.03 mm wide, 0.17 mm high (greatest height); III 3.03× 50 51 holotype and paratype of B. magnifica sp.n. were preserved in the length of flagellomere II but with similar morphology, 51 52 one large piece of amber, which was divided into several pieces 3.57 mm long, 0.02 mm wide, 0.10 mm high basally, 0.04 mm 52 53 to allow study of the 48 arthropods in it, including ten additional high distally, 0.64× the antennal length; IV short, 0.19 mm long, 53 54 dipterans, mainly small brachycerans. 0.02 mm wide, 0.04 mm high. Flagellomere II with dorsal sur- 54 55 face straight. Flagellomere III with abundant, fine, subcircular 55 56 Etymology. Named after El Soplao, the amber locality where foveae (probably sensilla) c.12μm wide (greatest width), iden- 56 57 the holotype was found. tical to those on same flagellomere in B. magnifica, occurring 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 Fig. 5. Anatomical features of the holotype (CES-349.1) of Buccinatormyia magnifica gen. et sp.n. from El Soplao amber. (A) Head in dorsal view 45 as preserved; (B) reconstructed labellum in dorsal (left) and lateral (right) views; (C) reconstructed palp in dorsal (top) and lateral (bottom) views; (D) 46 46 thorax in lateral view. 47 47 48 48 49 on the proximal third but without apparent pattern; this flag- Thorax: Only a small portion of the dorsal cuticle is preserved, 49 50 ellomere with a narrow ventrolateral band comprising dense, having long setae (such setae absent in B. magnifica). 50 51 abundant, minute, elongate foveae (7 μm long) (also probably Wing: Incomplete [small basal portion and part of anal lobe 51 52 sensilla, but about half the size of other foveae). Flagellomeres not preserved (cell cup incomplete)]. Length 9.07 mm, 3.10 mm 52 53 II and III broadly articulated with proximal article, especially wide (estimated greatest width) (W/L = 0.34). Costal vein end- 53 54 flagellomere II, indicating significant lateral mobility. Flagel- ing between R5 and M1. Veins Sc and R fused basal to crossvein 54 55 lomeres III and IV acuminate distally, but III only on ventral h; Sc 4.46 mm long (0.49× the wing length). Vein Sc meets 55 56 margin. Minute apical stylus present, slightly longer than in wing margin beyond middle of wing length, beyond level of 56

57 B. magnifica. r-m crossvein. Vein R1 5.26 mm long; stem of Rs relatively 57

1 1 veins CuA1 + M3 and CuA2 + A1 not preserved. Cell br nar- 2 rower than bm; cell bm slightly longer, 4.20 mm long (estimated) 2 3 vs. 4.04 mm and distally broader due to vein M arched costad. 3 4 Vein CuA with base approximately half the thickness of base 4 5 of R; CuP weakly developed, not sclerotized, parallel to CuA 5 6 6 and separated by distance equal to vein thickness. Veins A1 and 7 7 A2 not preserved. Alula and calypter not preserved. Wing mem- 8 brane entirely hyaline (no patterning or pterostigmatic sclero- 8 9 tization), evenly covered with minute microtrichia. Halters not 9 10 preserved. 10

11 Color 11 12 12 Linguatormyia Grimaldi, gen.n. 13 13 http://zoobank.org/urn:lsid:zoobank.org:act:B279141E-9204- 14 14 4B56-9504-16E7F43D7663 15 15 16 16 17 Type species. Linguatormyia teletacta gen. et sp.n. 17 18 18 19 Etymology. Derived from lingua- (L.), tongue, and -myia 19 20 (Gr.), fly, in reference to the long proboscis. 20 21 21 22 22 Diagnosis. Antenna with five articles (three flagellomeres), 23 23 without minute apical stylus; stem of Rs short (shorter than r-m 24 24 crossvein), M and M diverging directly from cell d (not from 25 1 2 25 a portion of M distad to cell d), vein A absent. 26 1+2 2 26 27 27 28 Linguatormyia teletacta Grimaldi, sp.n. 28 29 http://zoobank.org/urn:lsid:zoobank.org:act:8A386070-4F23- 29 30 45F9-A6A6-DD527B6A55BC 30 31 (Figs 10–13) 31 32 32 33 Type material. Holotype AMNH Bu-SE13, in Burmese 33 34 amber (Myanmar). The specimen is in a 36 × 30 × 11 mm piece 34 35 of dark, turbid amber containing fine particulate debris, trans- 35 36 verse fractures and flow lines (especially near the centre). These 36 37 obscure some features of the specimen, especially dorsally. The 37 38 entire head, palp, proboscis and even antennae are preserved, as 38 39 Fig. 6. Reconstruction of Buccinatormyia magnifica gen. et sp.n. in is the thorax and most of the abdomen (except for its tip). Lost 39 40 dorsal and lateral views to show its life aspect based on the holotype and at the surface of the amber are tip of the right wing and apices 40 paratype. The colour patterning is conjectural but based on the common 41 41 black-and-yellow patterns found in Xylomyidae and Stratiomyidae of all tibiae and all tarsi. Distal third of right wing is folded 3× 42 (Artist: J. A. Peñas). and required careful reconstruction. 42 43 43 44 Syninclusions. Two thrips (Thysanoptera), a small beetle, 44 long, 1.55 mm long (c. 5.5× the length of r-m crossvein); R 45 2+3 an empidoid fly, a portion of a neuropteran wing, and insect 45 2.85 mm long; stem of R 2.33 mm long, R and R short 46 4+5 4 5 frass. 46 47 but relatively longer in comparison to B. magnifica, asymmet- 47

48 rical, R4 1.19 mm long, R5 1.55 mm long; tip of R5 preapical 48 49 to wing tip. Wing tip less acute than in B. magnifica.Celld Etymology. Combination of tele (G.),far,andtactus (L.), 49 50 short and thick, 1.50 mm long (greatest length), 0.70 mm wide sense of touch, meaning ‘far touching’, in reference to the 50 strikingly elongate antennae. 51 (W/L = 0.47). Stem of M1+2 1.17 mm long; M1 arched costad, 51 52 tip of M1 ending posterior to wing tip; M1 2.02 mm long (linear, 52 53 not along curvature), M2 1.42 mm long; M1 and M2 virtually Diagnosis. As for genus, with the following additions: 53 54 diverge directly from cell d, not from a portion of M1+2 dis- proboscis very long (about 0.45× the body length); antenna 54 55 tadtocelld;veinsM2 and M3 highly divergent. Vein M2 ends flagellate and extremely elongate (about× 1.80 the wing 55 56 before wing margin. Cell m3 long and broad (small distal portion length), flagellomere III about 0.80× the entire antennal 56 57 lost), 1.66 mm long (estimated), 0.57 mm wide (W/L = 0.34); length. 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10

11 Color 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 50 50 Fig. 7. Anatomical features of the holotype of Buccinatormyia magnifica gen. et sp.n. from El Soplao amber (CES-349.1), except subfigures (F and 51 G) from the paratype (CES-015.1/CES-392.3). (A, B) Labellum in dorsolateral and lateral views showing labellar lobes with fine pseudotracheae; 51 52 (C) metatibia showing the intersegmental membrane with two tibial spurs (arrows); (D) metatarsus showing paddle-shaped pulvilli, empodium and 52 53 claws covered by microtrichia; (E) pretarsus in lateral view; (F) wing (inset: sclerotized CuP); (G) halter in dorsal view; (H) two of the major abdominal 53 54 ‘lateral flanges’ with series of abdominal muscle plaques; (I) margin of an abdominal ‘lateral flange’ showing dense vestiture of ‘stellate hairs’(arrows), 54 55 constituted by microtrichia arranged in bunches and large, stiff deeply-ribbed setae (stellate in transverse section) similar to those on proboscis (asterisks); 55 56 (J) female terminalia with two-segmented cerci. Subfigures (C, D) and (I, J) made with consecutive pictures taken at successive focal planes.rs: Scaleba 56 57 (A–C, G, J) 0.2 mm; (D, E, I) 0.1 mm; (F) 1.0 mm; (H) 0.5 mm. 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 50 50 51 51 52 52 53 53 Fig. 8. Comparison between antennae (A) and wings (B) of Buccinatormyia magnifica gen. et sp.n. (top in both subfigures) and B. soplaensis sp.n. 54 54 (bottom in both subfigures), drawn to the same scale. Antennae depicted in fully lateral and dorsal views. Antennal viewsof B. magnifica enlarged for 55 relative comparison of antennal articles with those from B. soplaensis. Insets in antennal views of B. soplaensis showing fovea on flagellomere III (left 55 56 inset) and minute apical stylus (right inset). Wing of B. magnifica taken from the paratype; rest of the subfigures taken from their respective holotypes. 56 57 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10

11 Color 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 Fig. 9. Anatomical features of Buccinatormyia soplaensis sp.n. from El Soplao amber, holotype CES-015.2/CES-392.2. (A) Oblique section of the 43 44 proboscis showing the labrum; (B) detail of the same section showing the large, stiff deeply-ribbed setae with stellate transverse section; (C) cushion-like 44 45 labellar lobe showing abundant, fine pseudotracheae transversally crossed by abundant pseudotracheal rings; (D) broad articulation of flagellomere II; 45 46 (E) detail of the two groups of fine foveae on the proximal third of flagellomere III; (F) detail of the narrow ventral band of small foveae onthesame 46 47 flagellomere; (G) flagellomere IV with minute apical stylus; (H) wing (inset: weakly developed CuP). Subfigures (C, E and F) made withconsecutive 47 48 pictures taken at successive focal planes. Scale bars: (A, E) 0.1 mm; (B, F) 0.02 mm; (C, G) 0.2 mm; (D) 0.05 mm; (H) 1.00 mm. 48 49 49 50 Description (sex unknown). Body 8.92 mm long. antennal socket). Anterior ocellus apparently present, small; 50 51 51 Head: Broad, 2.67 mm wide. Eyes large, occupying nearly posterior ocelli either minute or lost. Vertex of head with para- 52 52 entire lateral surface of head (genae exposed ventrally), widely median emarginations, centrally raised into bare, median dor- 53 53 54 separated (ventrally 1.22 mm between inner margins); bare, sal tubercle; vertex with short, stiff setulae. Areas surrounding 54 55 devoid of interfacetal setulae; facets undifferentiated; inner mar- antennal bases raised into bare, rounded mounds (=facial lobes), 55 56 gins in frontal view with slight emarginations around anten- nearly meeting medially. Front of head with well-developed, 56 57 nal base (depth of emargination approx. half the diameter of inverted Y-shaped sulcus; diverging arms separating clypeus and 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10

11 Color 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 Fig. 10. Linguatormyia teletacta gen. et sp.n. in Burmese amber, holotype AMNH Bu-SE13. (A) Photomicrograph of amber piece containing the 40 40 specimen. Arrows indicate body of fly (bottom) and tips of antennae (top); (B) Drawing of entire holotype (ventral view), shown aspreserved. 41 41 42 42 43 ventral margins of facial lobes, these arms meet frontogenal tapered in width or depth apicad; covered with dense, thick 43 44 sulci beneath antennal sockets; straight arm running between microtrichia for entire length. Base of proboscis ventrally with 44 45 pair of facial lobes up frons to anterior ocellus. Clypeus bare, annulated membrane (ventral rostral membrane), ten annula- 45 46 raised into low mound between facial lobes and palpal bases, tions exposed (membrane probably telescoped). Dorsal portion 46 47 fully frontal in position; with short vertical, median suture; ven- of proboscis semi-tubular (labrum probably forming dorsal wall, 47 48 tral surface of clypeus membranous, oval, membrane connected epi/hypopharynx not visible, but most likely forming floor), sur- 48 49 to base of proboscis, with dorsomedial ligament. Frontoclypeal rounding hollow food canal; smaller ventral portion solid (prob- 49 50 sulci (demarcating lateral edges of clypeus) very deep, expos- ably a distended labium and/or prementum). Labellum very 50 51 ing pair of anterior tentorial pits. Genae developed into pair of small, as wide as proboscis, depth slightly greater than width; 51 52 rounded, ventral lobes; lobes slightly separated beneath base of length slightly less than depth; labellar lobes closely pressed 52 53 proboscis (not touching); with dense, fine setulae. Neck entirely together medially, with fine pseudotracheae on exposed sur- 53 54 membranous, 0.80 mm thick. faces. 54 55 Proboscis: Long (virtually all of it haustellum), 4.37 mm long Maxillary palp: Elongate, 1.47 mm long, projected for- 55 56 (measured from CT scan), 0.46× the body length. Proboscis lat- ward, two-segmented. Bases of palps flank base of proboscis 56 57 erally flattened (width 0.43× the cross-sectional depth); barely (as preserved, apical palpomeres virtually in contact with 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36

37 in Burmese amber, holotype AMNH Bu-SE13. (A) Fronto-ventral view; (B) ventral view; (C) oblique frontal view. 37 38 38 39 39 40 40

41 gen. et sp.n. 41 42 42 43 43 44 44 45 45 46 46 47 47

48 Linguatormyia teletacta 48 49 49 50 50 51 51 52 52 53 53

54 Nanoscans of head of 54 55 55 56 56

57 Fig. 11. 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 Fig. 12. Reconstructed habitus of Linguatormyia teletacta gen. et sp.n. in Burmese amber. 22 22 23 23 24 antennae). Basal palpomere long, 0.80× the total palp length; Immediately posterior to propresternum is probasisternum, c. 24 25 tubular, 0.20 mm wide; curved gently upward; ventral half 3–4× the size of former sclerite, with sulcus demarcating pro- 25 26 covered with fine, stiff setae; with very fine transverse suture furca well defined. Lateral and dorsal to cervical sclerites isa 26 27 (appears under transmitted visible light; this suture is not pair of flattened, triangular postpronotal lobes with short pilos- 27 28 visible in CT scans, indicating it may be only internally devel- ity. Between these lobes and probasisternum is small, trian- 28 29 oped). Apical palpomere short, oval, not pedunculate, laterally gular sclerite with bulging lobe, i.e. proanepisternum. Base of 29 30 flattened, pointed anteriad, bare; lateral surface with three proanepisternum articulating with base of procoxa. Mesotho- 30 31 longitudinal rows of fine foveae (probably sensilla). rax large, expansive. Katepisternum ventrally broad and slightly 31 32 Antenna: Five-articled (with three flagellomeres), project- bulging, with median sulcus of furca well defined. Anepister- 32 33 ing forward, extremely elongate and flagellate; 15.40 mm long num very large; anepimeron smaller, subdivided by partial cleft; 33 34 (1.62× estimated body length) (as preserved antennae converge katatergite rounded, bulging, setulose; mesothoracic spiracle 34 35 and cross twice, at 0.5× and 0.65× their total length). Anten- immediately ventral to katatergite; meron slightly larger than 35 36 nal bases broadly separated by distance 3× diameter of antennal mesocoxa. Coxae and ventral surfaces of femora with sparse, 36 37 socket (1.75× distance between palp bases), lying within shal- short setulae. Procoxae slender, 1.01 mm long, well separated 37 38 low emarginations of inner margin of eyes. Scape short, slightly from each other; mesocoxae closer together, bases nearly con- 38 39 wider than long; as preserved both scapes are oriented anterolat- tiguous, 0.58 mm long, 0.51 mm wide basally; metacoxal bases 39 40 erally (divergent). Pedicel short, 0.37× length of scape, 0.48× contiguous, 0.45 mm long, anterior surface of metacoxa with a 40 41 the width [pedicels and flagellomeres I pointed mesally (con- small, protruding lobe. Trochanters small, slender, fully artic- 41 42 vergent)]. Flagellum extremely long, slender, tapered, laterally ulated with femora and coxae (no fusion). Femora slender, 42 43 flattened; with dense, fine setulae. Flagellum comprising three profemur 2.14 mm long, mesofemur 2.12 mm long, metafemur 43 44 flagellomeres: I as long as scape, 0.40 mm long, 0.16 mm wide; 2.89 mm long; each dorsally with dense, recumbent setulae. Tib- 44 45 II 0.97 mm long, 0.13 mm wide; III 12.8 mm long, 0.83× the iae slender (none with apex preserved), entirely covered with 45 46 antennal length. Minute apical stylus absent. dense setulae. 46 47 Thorax: Broad and deep, 2.68 mm wide, slightly wider than Wing: Length 8.64 mm (from reconstruction); 3.39 mm wide 47 48 head; devoid of macrosetae, exocuticle with fine, short, pile-like (greatest width) (W/L = 0.39). Vein C not circumambient (wing 48 49 setae only; coloration appears uniformly dark brownish (pos- tip lost). Veins Sc and R fused basal to crossvein h; Sc 3.71 mm 49 50 sibly preservational). Prothorax highly reduced, as in other long (0.36× the wing length), Sc ending beyond level of r-m 50

51 Diptera, but with some modifications. Propresternum a small crossvein. Vein R1 4.82 mm long, gradually thickened apically; 51 52 central, anterior sclerite with pair of small lobes protrud- stem of Rs short, 0.50 mm long; R2+3 3.64 mm long; stem 52 53 ing anteriorly (lobe diameter 0.12 mm). Flanking these lobes of R4+5 3.43 mm long, R4 and R5 short, asymmetrical, R4 53 54 is pair of hemispherical, protruding lobes, diameter 0.25 mm 0.93 mm long, R5 1.34 mm long, tip of R5 preapical to wing tip. 54 55 (=cervical sclerites) shifted ventrally from lateral positions typ- Cell d short and thick, 1.50 mm greatest length, 0.68 mm wide 55

56 ical of most flies; cervical sclerite extremely heavily sclerotized (W/L = 0.45). Stem of M1+2 1.57 mm long; M1 arched costad, tip 56 57 (based on density in CT scans), with covering of sparse setulae. of M1 probably ending posterior to wing tip (based on trajectory 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 50 50 51 51 52 52 53 53 54 54 55 55 56 Fig. 13. Wings of recent (A–C) and fossil (D) stratiomyomorphans. (A) Xylomyia americana (Xylomyidae); (B) Allognosta fuscitarsis (Stratiomyi- 56 dae); (C) Pantophthalmus sp. (Pantophthalmidae); (D) Linguatormyia teletacta gen. et sp.n. (Zhangsolvidae). 57 57

1 1 of vein), M1 2.25 mm long (linear, not along curvature); M2 pedunculate and with a broad, rounded apex. In contrast, in 2 2 1.86 mm long; M1 and M2 diverge directly from cell d, not L. teletacta the basal palpomere is very long (0.80× the total palp 3 3 from a portion of M1+2 distad to cell d; veins M2 and M3 length), tubular and curved gently upward, whereas the apical 4 4 nearly parallel. Apices of veins R4,R5,M1 and M2 very thin, palpomere is short, oval, not pedunculate, and pointed anteriorly. 5 faint. Cell m3 long and slender, 1.96 mm long, 0.39 mm wide Moreover, the dorsal width of the eyes is broader in B. magnifica 5 6 (W/L = 0.20); veins CuA1 and M3 joined in very short vein (a condition apparently present as well in Cratomyia, though 6 7 before meeting wing margin. Cell br slightly narrower than bm; preservation in limestone makes this uncertain) but nearly equal 7 8 cell bm slightly longer (3.35 mm long vs. 3.00). Vein CuA with to the ventral width in L. teletacta (only a small portion of the 8 9 base very thick, approximately same thickness as base of R; eyes is preserved in B. soplaensis). 9

10 CuA2 fusing with A1 prior to wing margin, forming closed, Even though the palps, as mentioned, are not preserved in the 10 11 narrow cell cup; cup 4.42 mm long (W/L = 0.16); CuP well unique specimen of B. soplaensis, and that this species shares 11 12 developed, parallel to CuA and separated by distance equal to with L. teletacta the remarkable presence of flagellate antennae 12

13 vein thickness. Vein A1 entirely straight; no trace of A2 evident. – a unique feature among all extinct or extant Brachycera – 13 14 Anal lobe, alula and calypter well developed; calypter with we are placing the former within the genus Buccinatormyia 14 15 marginal vein. Wing membrane entirely hyaline (no patterning based on wing characters and antennal structure. Indeed, along 15 16 or pterostigmatic sclerotization), evenly covered with minute with the fewer antennomeres and the absence of an apical 16 17 microtrichia. Halter with slender apical club. stylus, the antenna of L. teletacta is much more elongate, 17 18 Abdomen: Broad, 3.00 mm wide, slightly broader than thorax, nearly 2× the wing length (vs. less than the wing length 18 19 as preserved dorsoventrally flattened. Five basal sternites and in B. soplaensis) and flagellomere III comprises 0.83× the 19 20 five basal tergites preserved, all with dense, short setulae. Ster- entire antennal length (vs. 0.64× in B. soplaensis). Zhangsolva 20 21 nites cover entire transverse width of abdomen; no abdominal possesses a basic antennal structure with multiple articles 21 22 pleural membrane exposed. Terminalia not preserved. having little differentiation (probably with no more than eight 22 23 flagellomeres; see comments above). 23 24 Remarks. Sex of the unique specimen cannot be defini- The presence of two morphotypes of a rare family, Zhang- 24 25 tively determined without terminalia, but the dichoptic con- solvidae, in the same piece of Spanish amber with similar 25 26 dition suggests it may actually be a female. Orthorrhaphous wing venation could be interpreted as a male and female of 26 27 Brachycera commonly have holoptic males, exceptions being the same species, rather than separate species as we propose. 27 28 all Apioceridae, Asilidae, Vermileonidaeand Xylomyidae, some Under this hypothesis, the three specimens were captured pos- 28 29 Xylophagidae and Bombyliidae, and the genus Tongamyia sibly during swarming or mating. However, B. magnifica (the 29 30 (Mydidae). smaller species) is represented by a female that has the ter- 30 31 Linguatormyia teletacta gen. et sp.n. differs from other minalia well preserved, and this size dimorphism would be 31 32 zhangsolvids in wing venation (most comparable structure for unlike any seen in orthorrhaphous Brachycera. Based on our 32 33 all fossil species) as follows: from Buccinatormyia soplaensis measurements of 30 exemplar genera and 34 species in the 33 34 three extant families of Stratiomyomorpha, males are usually 34 sp.n. by having much more slender cell m3 (W/L = 0.20, vs. 35 smaller than females, sometimes substantially so. The pres- 35 0.38), M2 and M3 nearly parallel (vs. highly divergent); from 36 ence and degree of body size dimorphism varies to some extent 36 Cratomyia by lacking a stem of M1 and M2 and by having 37 based on whether the measurement is thorax or wing length, 37 a broader wing (W/L = 0.39, vs. 0.35), lacking A2, and cell d 38 broader (W/L = 0.43, vs. 0.37); from Zhangsolva by the broader because in 18 of 26 stratiomyid species the females were larger 38 39 wing (vs. 0.34) and proximal end of cell d with two shallow, based on thorax length, and in 21 of 26 species females were 39 40 unequal sides (vs. equilateral ‘V’-shaped sides); and from the larger based on wing length (males were consistently smaller 40 41 other four genera by the short stem of Rs, which is shorter than in pantophthalmids and xylomyids). For all three families the 41 42 r-m crossvein (vs. several times the crossvein length). pattern of larger females is highly significantP ( ≪ 0.01, one- 42 43 and two-tailed sign tests). On average, female stratiomyids 43 44 were larger by 6–9% (depending on the length measurement), 44 45 Discussion female pantophthalmids by 6–20%, and female xylomyids by 45 46 9–16%. The wing length of the female holotype of B. magnifica 46 47 Some important characters have not been included in the is 6.00 mm – that of B. soplaensis is 9.07 mm (sex unknown). 47 48 diagnoses of the new genera because they are unknown from Therefore, if B. soplaensis were the male of B. magnifica its 44% 48 49 B. soplaensis sp.n. However, these characters are remarkably larger size would completely contradict the size dimorphism 49 50 different between B. magnifica gen. et sp.n. and Linguatormyia seen in extant stratiomyomorphans. 50 51 teletacta gen. et sp.n. and hence they hold a potential diagnostic Sexual dimorphisms of discrete characters in extant flies 51 52 significance between the two genera in which they have been can be very dramatic, commonly involving a character system 52 53 classified. The most important characters of these are basedon such as leg ornamentation, eye stalks, or body/wing coloration. 53 54 the palp morphology. In B. magnifica, the basal palpomere is Not only is the holotype of B. soplaensis considerably larger 54 55 long (0.60× the total palp length), inflated basally and gradually than the known female of the genus, but it has very long, 55 56 narrowing distally (with distal 2/3 strongly flattened laterally), flagellate antennae and different wing venation (see diagnosis 56 57 and straight; whereas the apical palpomere is long, club-like, of B. soplaensis, above). Sexual dimorphism in venation is rare 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10

11 Color 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 Fig. 14. Heads of recent stratiomyomorphans, showing some of the characters used in the phylogenetic analysis. (A, B) Oblique frontal views; (C, D) 36 37 fronto-ventral views. (A) Pantophthalmus kerteszianus (Pantophthalmidae); (B) Xylomyia americana (Xylomyidae); (C) Pantophthalmus comptus;(D) 37 38 Chiromyza sp. (Stratiomyidae). 38 39 39 40 in flies, including extant Stratiomyomorpha, and is generally Stratiomyidae, those in males being slightly smaller. The 40 41 subtle. Val (1976) records slight sexual differences in the antenna of female pantophthalmids is actually larger than in 41 42 venation of a few species of Pantophthalmus Thunberg, but, males, so in both these groups the sexual differences are oppo- 42 43 43 given the rarity of these flies, it is uncertain as to how consistent site what would be expected if B. magnifica and B. soplaensis 44 44 are these differences. The most obvious example involves were sexes of the same species. 45 45 Pantophthalmus batesi Austen, in which the size of cell m The body size, antennal structure and venation of the two 46 3 46 is slightly smaller in the male (cell width of 0.31 relative to species of Buccinatormyia are substantially distinctive from 47 47 length, vs. 0.40 in the female). In B. soplaensis,cellm is each other, such that we are confident these are not two sexes 48 3 48 49 actually much broader (0.38) than in the known female of the of the same species. It seems astonishing that two species of this 49 50 genus (0.22). rarely fossilized fly were captured in the same piece of amber. 50 51 Dimorphism in antennal structure does occur in Stratiomy- Perhaps copious resin exuded in close proximity to the plant 51 52 omorpha, most obvious in, again, the genus Pantophthalmus, reproductive structures that these flies visited for nectar. 52 53 where the female has a larger, stout flagellum (of eight flag- The following key to the genera of Zhangsolvidae is based 53 54 ellomeres) and the male a smaller, conical one with the api- mainly on wing venation. Antennal structure profoundly differs 54 55 cal article being styliform and 0.5–2× the length of the basal among some groups, but unfortunately this is ambiguous in 55 56 article (Fig. 14; Val, 1976). Woodley (1995) reported dimor- Zhangsolva and for Cratomyia macrorrhyncha the antennae are 56 57 phism in the size of antenna of about eight genera of beridine incompletely known. 57

1 Key to the genera of Zhangsolvidae ocellar tubercle, concave on each side of tubercle (Figs 1 2 11A, 14A). 2 3 1. M1 and M2 diverging from a portion of M1+2 distad to 2. Facial structure: (0) face flat or slightly bulging; (1) face 3 4 celld...... Cratomyia with small to large, conical tubercle, sometimes horn-like 4 5 (two species: compressions, Aptian, Crato Formation, Brazil) and larger in females (Fig. 14A, C). 5 6 –M1 and M2 divergingdirectlyfromcelld...... 2 3. Inner margin of eye around or just dorsal to base of 6 7 antenna: (0) no emargination; (1) slight emargination 7 2. Rs diverging from R1 significantly basal to the middle of the 8 wing. Crossvein r-m about 3× shorter than basal section of vein present (Figs 11A, 14B). 8 9 4. Pilosity of frons: (0) frons bare or with sparse pilosity; (1) 9 M1+2. Cell d about 3× longer than wide ...... Zhangsolva 10 (one species: compression, Barremian?, Laiyang Formation, frons with distinct pilosity (Fig. 14B). 10 11 China) 11 12 12 – Rs diverging from R1 near the middle of the wing. Antenna 13 Crossvein r-m as long as basal section of vein M or slightly 13 14 1+2 14 longer/shorter. Cell d <2.5× longer than wide ...... 3 5. Number of flagellomeres: (0) eight (Fig. 15A, B); (1) five 15 15 or four (Figs 5A, 8A, 10B); (2) two. 16 3. Antenna with six articles, with minute apical stylus. Stem of 16 6. Sexual dimorphism of apical flagellomere: (0) male simi- 17 Rs long, several times the r-m crossvein length...... 17 lar to female; (1) apical flagellomere slender and styliform 18 ...... Buccinatormyia 18 in males (Fig. 14C). 19 (two species: amber inclusions, Albian, El Soplao, Las Peñosas 19 7. Presence/absence and sensillar structure of foveae on the 20 Formation, Spain) 20 flagellum: (0) foveae absent; (1) foveae present and well 21 – Antenna with five articles, without minute apical stylus. Stem 21 developed with each fovea having a single, digitiform sen- 22 of Rs short, shorter than r-m crossvein ...... Linguatormyia 22 sillum basiconicum (Fig. 15); (2) foveae well developed, 23 (one species: amber inclusion, Upper Albian–Lower 23 with each sensillum basiconicum having multiple (≥2) 24 Cenomanian, Hukawng Valley, Myanmar) 24 digitiform processes (Fig. 15C). 25 25 8. Antennal stylus: (0) absent, or apical flagellomere not 26 Relationships and phylogenetic analysis 26 highly differentiated from preapical one; (1) present, 27 27 minute and conical (Figs 8A, 9G, 14A, 15A, D–G). 28 According to Mazzarolo & Amorim (2000), Zhangsolvidae 28 29 (their Cratomyiidae) is the sister group to (Xylomyi- 29 Mouthparts 30 dae + Stratiomyidae), with Pantophthalmidae the sister group 30 31 to these three families. That hypothesis was based on 11 char- 31 32 acters, all except one being wing venation. With Zhangsolvidae 9. Proboscis structure: (0) proboscis short, not extended 32 33 now known in amber, the preservation of critical, minute char- much beyond palps; (1) proboscis elongate, one to 33 34 acters allows an analysis of relationships of these interesting several times length of the head (Figs 1–6, 10–12). 34 35 flies with substantially more data. 10. Size of labellum: (0) labellum typical of most Brachyc- 35 36 A total of 52 morphological characters were scored for era, extended laterally approximately to palpal margins; 36 37 33 ingroup taxa (including 11 Cretaceous species) and two (1) labellum small, proboscis short in proportion to palps 37 38 outgroups (Symphoromyia sp. [Rhagionidae] and Xylophagus and head (Fig. 14C). 38 < 39 sp. [Xylophagidae]). Several larval characters are quite infor- 11. Palp structure: (0) palp short, length 5× the width; 39 40 mative of relationships, such as the presence of calcareous setose, but setae rarely like a bottle-brush; (1) palp long, 40 > 41 warts and three instars (Xylomyidae + Stratiomyidae) (Wood- length 10× width; with very long, erect setae at the 41 42 ley, 1986, 1995). However, we did not include immature char- base (Figs 5C, 11C). 42 43 acters because these are unknown for many extant and all fossil 12. Cuticular structure of palp: (0) without fine wrinkling; 43 44 taxa. Among the 52 characters, 15 are from the head (includ- (1) palpomeres with fine, irregular, transverse annula- 44 45 ing antennae and mouthparts), 10 from the thorax and legs, 15 tions/wrinkling at proximal ends (Fig. 14C). 45 46 from the wing venation, 5 from the pre-genital abdomen, and 13. Development of adult mouthparts (labellum, labium, 46 47 7 from the male and female terminalia. Woodley (1986, 1995), palps): (0) well developed, fully functional; (1) highly 47 48 Mazzarolo & Amorim (2000) and Yeates (2002) were sources reduced to vestigial (Fig. 14D). 48 49 for many of the characters, but 19 characters presented here are 49 50 newly proposed. Cratomyia, which is extremely well preserved Eyes 50 51 as mineralized 3D replicas in limestone, had twice the number 51 52 of missing characters as did the two amber genera. 14. Male eyes: (0) holoptic (Fig. 14C, D); (1) dichoptic 52 Characters used in the phylogenetic analysis are the following: 53 (Fig. 14B). 53 54 15. Eye vestiture: (0) absent, very sparse, or so minute as 54 Head 55 to be discernible under light microscopy at ca. 100× 55 56 1. Shape of vertex: (0) ocellar area only slightly raised, (Fig. 14A–D); (1) with dense, long pilosity (setal length 56 57 without lateral concavities; (1) head vertex with dorsal greater than diameter of eye facets). 57

1 1 Thorax 34. Thickness of base of vein M1: (0) vein well developed, 2 no thinner than other M veins (Figs 7F, 8B, 13A, D); 2 3 16. Post-tegula pilosity: (0) absent; (1) present. (1) vein base very thin to evanescent where it bisects the 3 4 17. Area surrounding metathoracic spiracle: (0) bare; (1) basal cell into br and bm cells (Fig. 13B). 4 5 5 with tuft of fine hairs ventrally (Fig. 16C). 35. Apex of vein M1: (0) distant from wing apex; (1) ending 6 18. Scutellar spines: (0) absent; (1) present. at or very near the wing apex (Fig. 13B); (2) evanescent 6 7 19. Bulla-like structure ventral to base of wing: (0) absent; or incomplete. 7 8 8 (1) present (Fig. 16C). 36. Divergence of veins M1 and M2: (0) directly from cell d 9 9 20. Probasisternum structure: (0) probasisternum of unmod- (not from a portion of M1+2 distad to cell d); (1) from a 10 10 ified size, distinct from prosternum and cervical scle- portion of M1+2 distad to cell d. 11 rites (Fig. 11B); (1) probasisternum large, fused with 37. Vein M3: (0) present; (1) absent (Fig. 13B). 11 12 surrounding sclerites. 38. Cell m3 shape (only applicable in those taxa where M3 12 13 is present): (0) cell open apically; (1) cell closed (i.e. M3 13 14 Legs and CuA1 meeting before wing margin) (Fig. 13A, C, 14 15 D). 15 16 21. Number of protibial spurs: (0) none; (1) one. 39. Proportions of discal cell: (0) cell long and narrow, 16 17 22. Number of mesotibial spurs: (0) none; (1) one (Fig. 16D, approximately 0.20× the wing length and greater 17 18 F); (2) two (Fig. 16E). (Fig. 13A); (1) cell short and broad (Fig. 13D). 18 19 23. Number of metatibial spurs: (0) none; (1) one; (2) two 40. Apical shape of cell cup: (0) acute (Figs 8B, 13A, C, D); 19

20 (Fig. 7C). (1) truncate (apical section CuA2 nearly perpendicular 20 21 24. Metafemur with lateroapical spine: (0) absent; to A1) (Fig. 13B). 21 22 (1) present. 22 23 25. Ornamentation of ventral surface of metafemur: (0) Abdomen 23 24 without spines or denticles; (1) with longitudinal series 24 25 of black denticles. 41. Structure of female cerci: (0) cerci not separated by 25 26 tergite X; (1) cerci separated by tergite X. 26 27 Wing 42. Tergal grooves: (0) absent; (1) present, deep and trans- 27 28 verse on tergites II–VI preapically (Fig. 16A, B). 28 29 26. Wing membrane: (0) without fine wrinkling; (1) with 43. Abdominal muscle plaques: (0) scattered, in an irregular 29 30 a series of fine wrinkles along the wing margin that line oblique to the tergal margins (Figs 7H, 16A); (1) 30 31 parallel the terminal sections of the veins; (2) with fine, parallel to the tergal sulci/margins (Fig. 16B). 31 32 irregular pleating/wrinkling that covers the entire wing; 44. Development of tergites I and II: (0) well developed (at 32 33 wrinkles without common orientation (Fig. 13C). least tergite II), as in tergites III–VI; (1) reduced, with 33 34 27. Vein C: (0) circumambient (Fig. 13C); (1) ending at or membranous areas; abdomen held raised. 34

35 slight past apex of wing (e.g. at apex of R5,M1,orM2) 45. Development of tergites VI and VII: (0) abdominal 35 36 (Fig. 8B); (2) ending before apex of wing, sometimes segments VI and VII slightly smaller than preceding 36

37 considerably so (e.g. at apex of R4,R4+5, or between segments, but not reduced; (1) abdominal segments VI 37 38 these and the wing tip) (Fig. 13B). and VII clearly reduced in size (i.e. five major segments 38 39 28. Basal branching of vein R: (0) Rs branching off R present). 39 40 basally, in proximal half of wing; (1) Rs branching off 46. Spermathecal number: (0) three; (1) two. 40

41 R distally (stem of Rs short), i.e. Rs diverging from R1 47. Shape of ejaculatory apodeme: (0) generally elongate, 41 42 near the middle of the wing. laterally compressed; (1) a concave plate. 42

43 29. Relative length of vein R1: (0) long, approximately 48. Sperm pump: (0) structure generalized; (1) uniquely 43 44 0.40× length of R (=base + R1) and higher (Fig. 13A, C, modified. 44 45 D); (1) short, from 0.15 to 0.30× length of R (Fig. 13B). 49. Hypandrial fusion: (0) hypandrium articulates with but 45

46 30. Orientation and structure of vein R4: (0) short, signifi- not fused to gonocoxites; (1) hypandrium fused to 46 47 cantly diverging from R5; (1) long, curved only at base gonocoxites. 47 48 (i.e. nearly parallel to R5). 50. Paramere structure: (0) parameres well developed, not 48 49 31. Apices of veins R4 and R5: (0) anterior to the wing tip fused or hood-like; (1) parameres fused, small, not 49 50 or R5 posterior to the wing tip (Figs 8B, 13B, D); (1) hood-like 50 51 encompassing the wing tip (Fig. 13C). 51. Size of male sternite VIII: (0) sternite a recognizable, 51 52 32. Costalization of radial veins (i.e. veins shortened, plate-like sclerite; (1) sternite reduced or absent. 52 53 crowded near anterior portion of costal vein): (0) 52. Phallus structure: (0) phallus a simple tube; (1) phallic 53 54 absent; (1) present (Fig. 13B). complex trifid. 54

55 33. Structure of vein M1: (0) vein barely curved or straight 55 56 (Fig. 13B); (1) vein slightly arched (Fig. 13A, C); (2) The TNT analysis of the matrix presented in Table 1 yielded 56 57 vein strongly arched (Figs 8B, 13D). 270 most-parsimonious trees of 101 steps. A consensus topology 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10

11 Color 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 50 50 51 51 52 52 53 Fig. 15. Antenna of recent basal Brachycera, showing details of the foveae and stylus. (A) Photomicrograph; (B–G) scanning electron micrographs. 53 54 (A) Artemita sp. (Stratiomyidae), lateral view of entire antenna; (B, C) Allognosta fuscitarsis (Stratiomyidae), showing entire antenna (B, in lateral view) 54 55 and detail of fovea with a multidigitate sensillum (C); (D) tip of antenna of Solva sp. (Xylomyidae), showing minute stylus; (E) tip of antenna of Xylomyia 55 56 pallidifemur (Xylomyidae), showing minute stylus; (F, G) tip of antenna of Xylophagus reflectans (Xylophagidae), showing apical flagellomere (F) and 56 57 detail of minute stylus (G). 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10

11 Color 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 Fig. 16. Thoracic and abdominal characters of recent Stratiomyomorpha. (A, B) Abdomens in dorsal view; (C) portion of thoracic pleura; (D–F) 29 30 mesotibial spurs (arrows). (A) Allognosta fuscitarsis (Stratiomyidae); (B) Pantophthalmus pictiventris (Pantophthalmidae); (C) Pantophthalmus 30 31 tabaninus;(D)Pantophthalmus frauenfeldi;(E)Solva sp. (Xylomyidae); (F) Allognosta fuscitarsis (Stratiomyidae). 31 32 32 33 is shown in Fig. 17 (CI 0.40, RI 0.76), and supports the of larval characters in our matrix. Also, our analysis did not 33 34 monophyly for each family: Pantophthalmidae, Xylomyidae, resolve basal relationships within Stratiomyidae. This is some- 34 35 Zhangsolvidae and Stratiomyidae. what surprising in light of the relationships among extant sub- 35 36 This analysis confirms Mazzarolo & Amorim’s (2000) place- families presented by Woodley (1986, 1995), of Parhadrestiinae 36 37 ment of Zhangsolvidae within the Stratiomyomorpha, a hypoth- (Chiromyzinae (Beridinae (Clade 4 [higher] Stratiomyidae))). 37

38 esis based on three wing venation characters: M3 fusedtoCuA1, Although we did recover monophyly of the Parhadrestiinae 38 39 forming a cell m3 (lost in many Stratiomyidae); CuA2 fused to (including Cretaceogaster), the Chiromyzinae and higher stra- 39 40 A1 (convergent with other families of Brachycera); and vein M1 tiomyids, we did not recover monophyly of the Beridinae. The 40 41 weakly to strongly arched (unique to stratiomyomorphans, but irresolution in our results is probably due to the inclusion of three 41 42 variable in development and not found in all). Though not known Early Cretaceous genera – i.e. Cretoxyla Grimaldi & Cumming 42 43 for all species of Zhangsolvidae, at least some of the amber (Xylomyidae, in Lebanese amber), Montsecia Mostovski (Stra- 43 44 species also share with extant stratiomyomorphans the following tiomyidae, in limestone from Spain) and Lysistrata Grimaldi 44 45 characters: (i) two palpomeres, (ii) no protibial spurs (observable & Arillo (Stratiomyidae, in Spanish amber) – as well as an 45 46 only for Buccinatormyia magnifica gen. et sp.n.) and (iii) vein undescribed taxon in Late Cretaceous amber from New Jer- 46AQ6 47 C ending near the wing apex instead of being circumambient sey (Grimaldi & Cumming, 1999). These fossils are incom- 47 48 (e.g. in Pantophthalmidae, which is putatively plesiomorphic but plete, basal, and have some unusual character combinations. The 48 49 could be structurally related to their large size and reinforcement monophyly of Beridinae is best defined by the presence ofa 49 50 of the wing). deep, transverse furrow on each of the larger tergites; Pantoph- 50 51 Interestingly, in our analysis the relationships among fami- thalmidae have similar, but more shallow, furrows, as do some 51 52 lies in the infraorder and to Xylophagus (Xylophagidae) were stratiomyids in Cretaceous amber that are as yet undescribed (D. 52 53 unresolved; the most widely accepted hypothesis of stra- Grimaldi, unpublished data). 53 54 tiomyomorphan relationships is Pantophthalmidae (Xylomyi- The analysis does indicate that there are three groups 54 55 dae + Stratiomyidae) (Woodley, 1986, 1995; Sinclair, 1992; in the Zhangsolvidae: Zhangsolva, Cratomyia and (Lin- 55 56 Sinclair et al., 1994; Yeates, 2002; Woodley et al., 2009; Wieg- guatormyia + Buccinatormyia), the latter defined by the 56 57 mann et al., 2011). This is undoubtedly due to the absence reduced number of flagellomeres;Cratomyia ( cretacica + 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 46 47 47 48 48 49 49 50 50 1100010101010010101002000201001010000100011000110000 1100010101010000101002010201001010000100011000110000 1100010101010000101002010201001010000100011000110000 0011001100000100000101200011000010000100000001110000 0011001100000100000101200011000010000100000001110000 0000201000000010000001000021010001101- 11100000110000 00??0??0000?0?0??0??02100011110011001- 00?0?00??????? 0000201000000000000001000021100001101- 11?0000?????0? ????0??00????????0?????00011110001001-1010?00??????? 0000100001001000000000000021100001101- 10100000111110 0000002000000000000001000021100001101- 11110000111111 0010002000000010010000000021100001101- 10110000111111 0010002000000100010000000021100001100010110000111111 0000002000000010010000000021100001101- 10110000111111 0000002000000010010001000021100001100010100010111111 0000002000000001010000000121100101201- 10100010111111 0000002000000101000000000121100101100010100010111111 0000002000000101000000000121100101200010100010111111 0000002000000001010000000121100101100010100010111111 0000002000000001000000000021100101100010100010111111 00100??0000?0?0000000??00010???0?1?01- 00?0000??????? ????0??110??0????0????????10000020000100????0??????? 10101?1110100?0000000220001100002000010000000??????? ????1?1110??0?0??????????011000020000100???????????? 10101?1010100?000000???00011000020000100?0000??????? ?0??0???10??0????0?0???0?011000020010100?0000??????? ?0??0???10??0????0?0???0?011000020010100?0000??????? 51 0000000001111111111222222222233333333334444444444555 1234567890123456789012345678901234567890123456789012 51

52 sp.0000200000000000000002200000010000000000000000000000 52 s 0000002000000101000000000121100101200010100010111111 53 0000001100000100000012200011000000000000000000000000 sp. 53 sp. 0000100001001000000000000021100001101- 10100000111110

54 Matrix for the phylogenetic analysis. 54 55 55 AB 0011001100000100000101201011000010000100000101110000 0011001100000100000101201011000010000100000101110000 56 56 Xylophagus Symphoromyia Tanapaulseni Nemotelusmeland Table 1. Opetiopsalienus Pantophthalmuspla Pantophthalmustab Xylomyiaamericana Xylomyiapallidif 00000??000000?000000???00011100001100010?000????????Solva Lysistrata Cretaceogaster 0010002000000101010000000021100101101- 10100010111111 NJamberstrat Montsecia Chiromyza Allognostafusci Berisfuscipes Exairetaspinigera Heteracantharufic Exodonthaluteipes Pachygastrinaesp Odontomyiapubesc Chrysochlorinapu Merosarguscingula Cyphomyiaalbitars Hermetiaillucen Cretoxylaazari Zhangsolvacupres Buccinatmagnific Buccinatsoplaensis Linguatormyiatele Cratomyiamacrorr Cratomyiacretac Outgroups: 57 Solva Parhadrestia 57

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 Fig. 17. Consensus cladogram of relation- 31 ships of recent and fossil Stratiomyomorpha 32 32 based on morphological characters (see text). 33 Outgroups were Symphoromyia sp. (Rhagion- 33 34 idae) and Xylophagus sp. (Xylophagidae). Taxa 34 35 marked by (†) are fossils from the Cretaceous. 35 36 36 37 C. macrorrhyncha) are grouped based on the distinctive stem Rafael López del Valle for preparing the Spanish specimens and 37 38 38 of M1+2 beyond the cell d. Although we were not able to deter- to José Antonio Peñas for making the figure reconstructions. 39 mine a close relationship of Zhangsolvidae to any family of Thanks are due to Sieghard Ellenberger, Kassel (Germany), 39 40 Stratiomyomorpha, the analysis further substantiates placement for providing the Burmese amber specimen for purchase to the 40 41 in this infraorder. The monophyly of Zhangsolvidae is defined AMNH; to Steve Thurston, AMNH, for Burmese amber graph- 41 42 by several features, the most striking being the unique devel- ics. This study is supported by the IGME project 491-CANOA 42 43 opment within Stratiomyomorpha of an elongate proboscis. A 35015, the Spanish Ministry of Economy and Competitive- 43 44 similar proboscis has developed convergently in assorted orth- ness project CGL2011-23948/BTE, the National Science 44 45 orrhaphous Brachycera, i.e. within Acroceridae, Apioceridae, Foundation grant 1305066, and by generous funding to the 45 46 Bombyliidae, Mydidae, Nemestrinidae, Tabanidae and Ver- American Museum of Natural History by Robert G. Goelet, 46 47 mileonidae, where it is always associated with flower-visiting Chairman Emeritus of the Board of Trustees. We are grateful 47 48 behaviour. As such, the proboscis in Zhangsolvidae has sig- for the expert reviews by two anonymous reviewers, which 48 49 nificant evolutionary implications, discussed elsewhere (E. significantly improved the original manuscript. 49 50 Peñalver et al., in preparation). 50 51 51 52 References 52 53 53 Acknowledgements Borkent, A. & Grimaldi, D.A. (2004) The earliest fossil mosquito 54 (Diptera: Culicidae), in Mid-Cretaceous Burmese amber. Annals of 54 55 We thank the staff of the El Soplao cave in Cantabria (Spain) the Entomological Society of America, 97, 882–888. 55 56 for curation and access to the Spanish specimens, the enterprise Goloboff, P.A., Farris, J.S. & Nixon, K.C. (2008) TNT, a free program 56 57 SIEC S.A. and the Government of Cantabria. We are grateful to for phylogenetic analysis. Cladistics, 24, 774–786. 57

1 Grimaldi, D.A. & Arillo, A. (2008) The Tethepomiidae, a new family of Shi, G., Grimaldi, D.A., Harlowb, G.E. et al. (2012) Age constraint 1 2 enigmatic Cretaceous Diptera. Alavesia, 2, 259–265. on Burmese amber based on U-Pb dating of zircons. Cretaceous 2 3 Grimaldi, D.A. & Cumming, J.M. (1999) Brachyceran Diptera in Research, 37, 155–163. 3 4 Cretaceous ambers and Mesozoic diversification of the Eremoneura. Sinclair, B.J. (1992) A phylogenetic interpretation of the Brachycera 4 5 Bulletin of the American Museum of Natural History, 239, 1–124. (Diptera) based on the larval mandible and associated mouthpart 5 Grimaldi, D.A., Engel, M.S. & Nascimbene, P.C. (2002) Fossiliferous structures. Systematic Entomology, 17, 233–252. 6 6 Cretaceous amber from Myanmar (Burma): its rediscovery, biotic Sinclair, B.J., Cumming, J.M. & Wood, D.M. (1994) Homology and 7 diversity, and paleontological significance. American Museum Novi- phylogenetic implications of male genitalia in Diptera – Lower 7 8 tates, 3361, 71 pp. Brachycera. Entomologica Scandinavica, 24, 407–432. 8 9 Grimaldi, D.A., Arillo, A., Cumming, J.M. & Hauser, M. (2011) Stoffolano, J.G., Woodley, N.E., Borkent, A. & Yin, L.R.S. (1988) 9 10 Brachyceran Diptera (Insecta) in Cretaceous ambers, part IV, signifi- Ultrastructural studies of the abdominal plaques of some Diptera. 10 11 cant new orthorrhaphous taxa. Zookeys, 148, 293–332. Annals of the Entomological Society of America, 81, 503–510. 11 12 Krzeminski,´ W. & Krzeminska,´ E. (2003) Triassic Diptera: descriptions, Stuckenberg, B.R. (2000) Namamyia, a new genus of Vermileonidae 12 13 revisions and phylogenetic relations. Acta Zoologica Cracoviensia, 46 (Diptera) from Namaqualand, with a conspectus of the Southern 13 (suppl.–Fossil Insects), 153–184. African genera and an account of their adaptations for anthophily. 14 14 Mazzarolo, L.A. & Amorim, D.S. (2000) Cratomyia macrorrhyncha, Annals of the Natal Museum, 41, 181–202. 15 a Lower Cretaceous brachyceran fossil from the Santana Forma- Szucsich, N.U. & Krenn, H.W. (2000) Morphology and function of the 15 16 tion, Brazil, representing a new species, genus and family of the proboscis in Bombyliidae (Diptera, Brachycera) and implications for 16 17 Stratiomyomorpha (Diptera). Insect Systematics & Evolution, 31, proboscis evolution in Brachycera. Zoomorphology, 120, 79–90. 17 18 91–102. Szucsich, N.U. & Krenn, H.W. (2002) Flies and concealed nectar 18 19 McAlpine, J.F. (1981). Morphology and terminology – Adults. Manual sources: morphological innovations in the proboscis of Bombyliidae 19 20 of Nearctic Diptera, Agriculture Canada Monographs, Vol. I (ed. (Diptera). Acta Zoologica (Stockholm), 83, 183–192. 20 21 by J.F. McAlpine, B.V. Peterson, G.E. Shewell, H.J. Teskey, J.R. Val, F.C. (1976) Systematics and evolution of the Panthopthalmi- 21 Vockeroth and D.M. Wood) 9–63. Research Branch, Ottawa. dae (Diptera, Brachycera). Arquivos de Zoologia (Sao Paulo), 27, 22AQ7 22 Nagatomi, A. & Yang, D. (1998) A review of extinct Mesozoic genera 51–164. 23 and families of Brachycera (Insecta, Diptera, Orthorrhapha). The Wiegmann, B.M., Trautwin, M.D., Winkler, I.S. et al. (2011) Episodic 23 24 Entomologists’ Monthly Magazine, 134, 95–192. radiations in the fly tree of life. Proceedings of the National Academy 24 25 Najarro, M., Peñalver, E., Rosales, I., Pérez-de la Fuente, R., of Sciences of the United States of America, 108, 5690–5695. 25 26 Daviero-Gomez, V., Gomez, B. & Delclòs, X. (2009) Unusual Wilkommen, J. & Grimaldi, D.A. (2007). 11.20 Diptera: true flies, gnats, 26 27 concentration of Early Albian arthropod-bearing amber in the and crane flies, The Crato Fossil Beds of Brazil (ed. by D.M. Martill, 27 28 Basque-Cantabrian Basin (El Soplao, Cantabria, Northern Spain): G. Bechly and R.F. Loveridge, pp. 369–387. Cambridge University 28 29 palaeoenvironmental and palaeobiological implications. Geologica Press. 29AQ9 Acta, 7, 363–387. Woodley, N.E. (1986) Parhadrestiinae, a new subfamily for Parhadestria 30 30 Najarro, M., Peñalver, E., Pérez-de la Fuente, R. et al. (2010) A James and Cretaceogaster Teskey (Diptera: Stratiomyidae). System- 31 review of the El Soplao amber outcrop, early Cretaceous of atic Entomology, 11, 377–387. 31 32 Cantabria (Spain). Acta Geologica Sinica (English Edition), 84, Woodley, N.E. (1995) The genera of Beridinae (Diptera: Stratiomyidae). 32 33 959–976. Memoirs of the Entomological Society of Washington, 16, 1–231. 33 34 Nascimbene, P. & Silverstein, H. (2000). The preparation of fragile Cre- Woodley, N.E. (2001) A World Catalog of the Stratiomyidae (Insecta: 34 35 taceous ambers for conservation and study of organismal inclusions. Diptera), Myia, Vol. 11. Backhuys Publishers, Leiden. 35AQ10 36 Studies on Fossils in Amber, with Particular Reference to the Cre- Woodley, N.E. (2011) A catalog of the world Xylomyidae (Insecta: 36 37 taceous of New Jersey (ed. by D. Grimaldi), pp. 93–102. Backhuys Diptera). Myia, 12, 485–521. 37 Publishers, Leiden. Woodley, N.E., Borkent, A. & Wheeler, T.A. (2009). Phylogeny of the 38 38 Peñalver, E. & Delclòs, X. (2010). Spanish Amber. Biodiversity of Diptera. Manual of Central American Diptera,Vol.I.(ed.byB.V 39 Fossils in Amber from the Major World Deposits (ed. by D. Penney), Brown, A. Borkent, J.M. Cumming, D.M. Wood, N.E. Woodley and 39 40 pp. 236–270. Siri Scientific Press: Manchester. M. Zumbado), 79–94. National Research Council Press, Ottawa. 40 41 Pérez-de la Fuente, R. (2012) Paleobiología de los artrópodos del Yeates, D.K. (2002) Relationships of the lower Brachycera (Diptera): 41 42 ámbar cretácico de El Soplao (Cantabria, España). Doctoral Thesis, a quantitative synthesis of morphological characters. Zoologica 42 43 unpublished, Barcelona University, Barcelona [In Spanish, with Scripta, 31, 105–121. 43 44AQ8 English summary]. Zhang, J.F., Zhang, S. & Li, L.Y. (1993) Mesozoic gadflies (Insecta: 44 45 Pérez-de la Fuente, R., Delclòs, X., Peñalver, E. & Arillo, A. (2011) Diptera). Acta Palaeontologica Sinica, 26, 595–603. 45 Biting midges (Diptera: Ceratopogonidae) from the Early Cretaceous Zherikhin, V.V. & Ross, A.J. (2000) A review of the history, geology 46 46 El Soplao amber (N Spain). Cretaceous Research, 32, 750–761. and age of Burmese amber (Burmite). Bulletin of the Natural History 47 Ross, A.; Mellish, C.; York, P. & Crighton, B. (2010). Burmese Amber. Museum, London (Geology), 56, 3–10. 47 48 Biodiversity of Fossils in Amber from the Major World Deposits (ed. 48 49 by D. Penney), pp. 208–235. Siri Scientific Press, Manchester. Accepted 9 August 2014 49 50 50 51 51 52 52 53 53 54 54 55 55 56 56 57 57

IMPORTANT NOTE: Please mark your corrections and answers to these queries directly onto the proof at the relevant place.

Queries from the Copyeditor:

AQ1. Please confirm if the inserted state name abbreviations are fine for affiliations 3,5and6. AQ2. please insert manufacturer’s name and place of manufacture at first mention) AQ3. We have treated “Family Zhangsolvidae Nagatomi & Yang, 1998” as head level 3. Please confirm if this fine. AQ4. Wilkommen, 2007 has not been included in the Reference List, please supply full publication details. AQ5. We have removed the two rows here as it is identical to the two header rows given above. Please confirm if its fine. AQ6. Late: or should this be Upper? AQ7. We have inserted the city location and name of the publisher for Reference McAlpine, 1981. Please confirm if this is correct. AQ8. As per the journal style thesis references does not have page range. So we have deleted the page range for reference Pérez-de la Fuente, 2012. Please confirm if is this fine. AQ9. Please provide the location (city, state/country) of publisher name in the reference Wilkommen and Grimaldi, 2007. AQ10. is Myia a book series or journal title (or both, as in the two Woodley refs) AQ11. We have processed Figure(s) 2, 4, 6 and 7, 9 and 10, 14–16 in colour. If these figures are to be published in colour, please fill out CWA form if not yet done.